Modulo Indexing, Periodic Extension

GUIDE: Mathematics of the Discrete Fourier Transform (DFT). Modulo Indexing, Periodic Extension

It appears that you are using AdBlocking software. The cost of running this website is covered by advertisements. If you like it please feel free to a small amount of money to secure the future of this website.

<< Previous page  TOC  INDEX  Next page >>

Modulo Indexing, Periodic Extension

The DFT sinusoids $s_k(n) \ are all periodichaving periods which divide $N$. That is, $s_k(n+mN)=s_k(n)$ for any integer $m$. Since a length $N$ signal $x$ can be expressed as a linear combination of the DFT sinusoids in the time domain,

\

it follows that the “automatic” definition of $x(n)$ beyond the range$[0,N-1]$ is periodic extension, i.e., $x(n+mN)\ for every integer $m$.

Moreover, the DFT also repeats naturally every $N$ samples, since

\

because $s_{k+mN}(n) = e^{j2\. (The DFT sinusoids behave identically as functions of $n$ and$k$.) Accordingly, for purposes of DFT studies, we may define allsignals in ${\ as being single periods from an infinitely long periodic signal with period $N$ samples:



Definition: For any signal $x\, we define

\

for every integer $m$.

As a result of this convention, all indexing of signals and spectra8.1 can be interpreted modulo $N$, and we may write $x(n \ to emphasize this. Formally, “$n \” is defined as$n-mN$ with $m$ chosen to give $n-mN$ in the range $[0,N-1]$.

As an example, when indexing a spectrum $X$, we have that $X(N)=X(0)$ which can be interpreted physically as saying that the sampling rate is the samefrequency as dc for discrete time signals. In the time domain, we have what is sometimes called the “periodic extension” of $x(n)$. This means that the input to the DFT is mathematically treated as samples of a periodic signal with period $NT$ seconds ($N$ samples). The corresponding assumption in the frequency domain is that the spectrum is zero between frequency samples $\.

It is also possible to adopt the point of view that the time-domain signal$x(n)$ consists of $N$ samples preceded and followed by zeros. In this case, the spectrum is nonzero between spectral samples$\, and the spectrum between samples can be reconstructed by means of bandlimited interpolation. This “time-limited” interpretation of the DFT input is considered in detail in Music 420 and is beyond the scope of Music 320 (except in the discussion of “zero padding $\interpolation” below).

<< Previous page  TOC  INDEX  Next page >>

 

© 1998-2017 – Nicola Asuni - Tecnick.com - All rights reserved.
about - disclaimer - privacy